Papers by Xiangdong Su

25 papers
A Mutual Information Perspective on Knowledge Graph Embedding (2025.acl-long)

Copied to clipboard

Challenge: Existing knowledge graph embedding techniques suffer from high intra-group similarity, loss of semantic information, and insufficient inference capability, particularly in complex relation patterns such as 1-N and N-1 relations.
Approach: They propose a knowledge graph embedding framework that leverages mutual information maximization to improve the semantic representation of entities and relations.
Outcome: Extensive experiments on benchmark datasets demonstrate the effectiveness of the proposed method, with consistent performance improvements across various baseline models.
APOLLO: An Optimized Training Approach for Long-form Numerical Reasoning (2024.lrec-main)

Copied to clipboard

Challenge: Existing methods to generate reasoning programs that ignore the differences between facts treated all facts equally, leading to wrong punishment of programs that differed from the ground truth.
Approach: They propose an optimized training framework for long-form numerical reasoning that incorporates a number-aware negative sampling strategy and consistency-based reinforcement learning to increase execution accuracy.
Outcome: The proposed method improves the performance of long-form numerical reasoning on the FinQA and ConvFinQA leaderboards.
TransERR: Translation-based Knowledge Graph Embedding via Efficient Relation Rotation (2024.lrec-main)

Copied to clipboard

Challenge: Existing knowledge graph embedding models lack links between entities and relationships, which is a problem for knowledge graphs.
Approach: They propose a translation-based knowledge geraph embedding method via efficient relation rotation that rotates the head and tail entities with their corresponding unit quaternions.
Outcome: The proposed method can be used to embed knowledge graphs on 10 benchmark datasets with fewer parameters than the previous translation-based models.
Know Your Place: Diagnosing Implicit Social Adaptation Failures in Chinese Large Language Models (2026.acl-long)

Copied to clipboard

Challenge: Existing studies suggest that failures of large language models in social contexts are not due to limited linguistic competence, but to inappropriate recognition.
Approach: They propose a framework that decomposes social adaptation into three orthogonal dimensions and conduct controlled comparisons across multiple Chinese LLMs under implicit and explicit conditions.
Outcome: The proposed framework decomposes social adaptation into three orthogonal dimensions and conducts controlled comparisons across multiple Chinese LLMs under implicit and explicit conditions.
CausalityCheck: A Framework for Evaluating Causal Reasoning in Large Language Models (2026.findings-acl)

Copied to clipboard

Challenge: Existing evaluation methods fail to accurately reflect a model's causal reasoning capabilities.
Approach: They propose a tool to automatically generate causal reasoning checklists to assess the causal reasoning abilities of 18 large language models.
Outcome: The proposed tool assesses the causal reasoning abilities of 18 large language models.
Learning Low-dimensional Multi-domain Knowledge Graph Embedding via Dual Archimedean Spirals (2024.findings-acl)

Copied to clipboard

Challenge: Existing knowledge graph embedding methods make domain constraints on embeddable domains, leading to poor performance.
Approach: They propose a low-dimensional KGE model for multi-domain knowledge graphs that embeds domains and domains by regularization function.
Outcome: The proposed model can distinguish entities from domains by encoding the same relation on the same archimedean spiral.
McBE: A Multi-task Chinese Bias Evaluation Benchmark for Large Language Models (2025.findings-acl)

Copied to clipboard

Challenge: Existing datasets on bias evaluation for large language models focus on English and North American culture and are limited to one task.
Approach: They propose to evaluate Chinese language models' biases from multiple perspectives using a multi-task Chinese Bias Evaluation Benchmark.
Outcome: The proposed model covers 12, 82 subcategories and 5 evaluation tasks covering a wide range of categories and content diversity.
F²Bench: An Open-ended Fairness Evaluation Benchmark for LLMs with Factuality Considerations (2025.emnlp-main)

Copied to clipboard

Challenge: Existing fairness evaluation benchmarks for large language models rely on closed-ended evaluation formats that overlook factuality considerations rooted in historical, social, physiological, and cultural contexts.
Approach: They propose an open-ended fairness evaluation benchmark for large language models . they incorporate factuality considerations and multi-turn reasoning into the benchmark .
Outcome: The proposed benchmark incorporates factual grounding and text generation to better reflect the complexities of real-world model usage.
Exploring the Capability Boundaries of LLMs in Mastering of Chinese Chouxiang Language (2026.findings-acl)

Copied to clipboard

Challenge: Current state-of-the-art LLMs exhibit clear limitations on multiple tasks, while performing well on tasks that involve contextual semantic understanding.
Approach: They propose a mouse-based benchmark to evaluate LLMs' performance on NLP tasks involving Chouxiang Language.
Outcome: The proposed benchmark evaluates the performance of LLMs on six NLP tasks involving Chouxiang Language.
EpLSA: Synergy of Expert-prefix Mixtures and Task-Oriented Latent Space Adaptation for Diverse Generative Reasoning (2024.lrec-main)

Copied to clipboard

Challenge: Existing models for diverse generative reasoning struggle to generate multiple unique and plausible results.
Approach: They propose a model based on expert-prefix mixtures and task-oriented latent space adaptation for diverse generative reasoning.
Outcome: The proposed model outperforms existing models on three types of generative reasoning tasks.
TeAST: Temporal Knowledge Graph Embedding via Archimedean Spiral Timeline (2023.acl-long)

Copied to clipboard

Challenge: Existing temporal knowledge graph embedding models fuse temporal information into entities, limiting their effectiveness and potential applications.
Approach: They propose a temporal knowledge graph embedding model which encodes Temporal knowledge graphs via Archimedean Spiral Timeline.
Outcome: The proposed model outperforms existing TKGE methods in terms of relational consistency and interpretability.
Mitigating Heterogeneity among Factor Tensors via Lie Group Manifolds for Tensor Decomposition Based Temporal Knowledge Graph Embedding (2025.naacl-long)

Copied to clipboard

Challenge: Existing studies have highlighted the effectiveness of tensor decomposition methods in the Temporal Knowledge Graphs Embedding task.
Approach: They propose a method that maps factor tensors onto a unified smooth Lie group manifold to approximate homogeneous in tensian decomposition.
Outcome: The proposed method can be directly integrated into existing tensor decomposition based TKGE methods without introducing extra parameters.
Incorporating Inner-word and Out-word Features for Mongolian Morphological Segmentation (2020.coling-main)

Copied to clipboard

Challenge: Mongolian morphological segmentation is a crucial preprocessing step in many Mongolian related NLP applications.
Approach: They propose a neural network incorporating inner-word and out-word features for Mongolian morphological segmentation.
Outcome: The proposed network is compared with baselines and evaluates its performance.
Exploring the Synergy of Dual-path Encoder and Alignment Module for Better Graph-to-Text Generation (2024.lrec-main)

Copied to clipboard

Challenge: KG-to-text generation model lacks explicit graph-text alignment strategy due to discrepancy between textual and structure information.
Approach: They propose a synergetic knowledge graph-to-text model with a dual-path encoder, alignment module and guidance module to solve these problems.
Outcome: The proposed model achieves competitive performance on three benchmark datasets.
CEDAR: A Chinese Evaluation Dataset for Computational Argumentation (2026.acl-long)

Copied to clipboard

Challenge: Existing debate datasets neglect important labels for argument mining, generation, and evaluation.
Approach: They propose a Chinese Evaluation Dataset for Computational Argumentation that includes key arguments and key rhetorical figures, debater roles, modal words, debate results and transcripts.
Outcome: The proposed dataset covers 600 debates about 318 topics from Chinese debate competitions.
How Well Apply Simple MLP to Incomplete Utterance Rewriting? (2023.acl-short)

Copied to clipboard

Challenge: Incomplete utterance rewriting (IUR) aims to restore incomplete utterant with sufficient context information for comprehension.
Approach: They propose a method to restore incomplete utterances with sufficient context information . they employ only one-layer MLP architecture to mine latent semantic information based on joint utterations .
Outcome: The proposed method is superior to existing methods in quality and speed.
Hyperbolic Representations for Prompt Learning (2024.lrec-main)

Copied to clipboard

Challenge: Existing techniques to train only continuous prompts while freezing the language model have been developed.
Approach: They propose to use hyperbolic space to model hierarchical relationships between prompts and inputs . they use a Poincaré disk to capture the hierarchic relationship between prompt and input .
Outcome: The proposed approach reduces training time and storage for downstream tasks by reducing training costs.
Born Pragmatic, Trained to Hallucinate? Quantifying the Origins of Contextual Bias in LLMs via the PaCE Benchmark (2026.findings-acl)

Copied to clipboard

Challenge: Large language models excel at capturing communicative intent, but they have a side effect: pragmatic hallucination.
Approach: They propose a benchmark to quantify the impact of pragmatic hallucination on large language models . they propose RLHF and SFT to induce a strong tendency for pragmatic over-attribution .
Outcome: The proposed model outperforms existing models in predicting pragmatic hallucinations . the evaluations show that current alignment paradigms lack precise control over pragmatic boundaries .
Learning Continuous Temporal Dynamics on Symplectic Manifolds for Temporal Knowledge Graph Embedding (2026.findings-acl)

Copied to clipboard

Challenge: Existing methods for temporal knowledge graph embedding lack explicit structural constraints for continuous-time dynamics.
Approach: They propose a Temporal Knowledge Graph Embedding framework that embeds temporal dynamics into a symplectic phase space.
Outcome: The proposed framework achieves competitive performance with lower embedding dimensions.
Lightweight Haar Wavelet Subband Pruning for LLMs (2026.findings-acl)

Copied to clipboard

Challenge: Large language models (LLMs) have impressive performance but require computational and memory resources.
Approach: They propose a post-training framework that uses a Haar wavelet transform to prune weights.
Outcome: The proposed pruning framework reduces pruning time and computational costs by removing less important weights while preserving model architecture.
C3LRSO: A Chinese Corpus for Complex Logical Reasoning in Sentence Ordering (2025.coling-main)

Copied to clipboard

Challenge: Existing sentence ordering datasets for non-English languages are unavailable.
Approach: They propose a parameter-free sentence ordering dataset that provides genuinely unordered sentences without artificial segmentation cues.
Outcome: The proposed method outperforms existing methods on the sentence ordering task.
ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model (2026.findings-acl)

Copied to clipboard

Challenge: Theory of Mind (ToM) is widely regarded as central to effective persuasion, yet existing evaluations fail to capture the infer–apply loop that arises in real-world dialogue.
Approach: They propose a benchmark that conditions on the audience persona p and the Elaboration Likelihood Model (ELM) route r within persuasive conversations.
Outcome: The proposed model can model the interlocutor's mental states over multiple turns and adapt strategy and tone accordingly.
Leveraging 3D Gaussian for Temporal Knowledge Graph Embedding (2025.findings-emnlp)

Copied to clipboard

Challenge: Representation learning in knowledge graphs (KGs) has focused on static data, yet many real-world knowledge graph are inherently dynamic.
Approach: They propose a temporal embedding method inspired by 3D Gaussian Splatting where entities, relations, and timestamps are modeled as 3D gaussian distributions with learnable structured covariance.
Outcome: The proposed method outperforms state-of-the-art methods on three benchmark TKG datasets.
Who Wrote This Line? Evaluating the Detection of LLM-Generated Classical Chinese Poetry (2026.acl-long)

Copied to clipboard

Challenge: a recent study shows that large language models can generate text, but they can also fabricate large amounts of false or misleading content.
Approach: They propose a benchmark to detect LLM-generated classical Chinese poetry . they compare 12 different AI detectors to find out whether a poem is authored by AI .
Outcome: The proposed benchmark compared 12 AI detectors with a dataset of 30,664 Chinese poems . the results highlight the limitations of current Chinese text detectors .
Ensuring Safe and High-Quality Outputs: A Guideline Library Approach for Language Models (2024.naacl-long)

Copied to clipboard

Challenge: Guide-Align is a guideline-oriented approach to augment the safety and quality of Large Language Models.
Approach: They propose a guideline-oriented method to augment the safety and quality of large language models.
Outcome: The proposed method outperforms existing methods on three benchmarks and shows significant improvements in security and quality.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations